Understanding How to Use Open-Source Libraries for Differentially Private Statistics on Energy Metering Time Series

Ana C. P. Paixão, Breno R. da Silva, Rafael L. Silva, Filipe H. Cardoso, Alexandre Braga

2025

Abstract

Demand forecasting and dynamic pricing for renewable energy open markets may require heavy analytics capabilities on fine-grained consumption data. With differential privacy, data aggregators in the energy sector can compute statistics on metering information without accidentally leaking consumption patterns of specific consumers over time. However, differential privacy is complex and hard to implement correctly. In this paper, we propose a method for evaluating differential privacy libraries by their ability to produce private and useful statistics on time series for energy consumption. The method was validated by applying it to three open source libraries used to compute differentially private averages, counts, and sums on energy metering data. The method was able to clearly distinguish between private (indistinguishable) and disclosed (distinguishable) statistics. Our method and findings can help data scientists and privacy officers within the energy sector better understand how open-source differential privacy libraries behave with time series for energy metering data.

Paper Citation

in Harvard Style

Paixão A., R. da Silva B., Silva R., Cardoso F. and Braga A. (2025). Understanding How to Use Open-Source Libraries for Differentially Private Statistics on Energy Metering Time Series. In Proceedings of the 10th International Conference on Internet of Things, Big Data and Security - Volume 1: IoTBDS; ISBN 978-989-758-750-4, SciTePress, pages 289-296. DOI: 10.5220/0013338500003944

in Bibtex Style

@conference{iotbds25,
author={Ana Paixão and Breno R. da Silva and Rafael Silva and Filipe Cardoso and Alexandre Braga},
title={Understanding How to Use Open-Source Libraries for Differentially Private Statistics on Energy Metering Time Series},
booktitle={Proceedings of the 10th International Conference on Internet of Things, Big Data and Security - Volume 1: IoTBDS},
year={2025},
pages={289-296},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0013338500003944},
isbn={978-989-758-750-4},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 10th International Conference on Internet of Things, Big Data and Security - Volume 1: IoTBDS
TI - Understanding How to Use Open-Source Libraries for Differentially Private Statistics on Energy Metering Time Series
SN - 978-989-758-750-4
AU - Paixão A.
AU - R. da Silva B.
AU - Silva R.
AU - Cardoso F.
AU - Braga A.
PY - 2025
SP - 289
EP - 296
DO - 10.5220/0013338500003944
PB - SciTePress